Projected Freshening of the Arctic Ocean in the 21st Century

Abstract

Using state‐of‐the‐art models from the Coupled Model Intercomparison Project phase 5 (CMIP5), this study found the upper Arctic Ocean likely to freshen considerably in the future. Arctic Ocean average sea surface salinity is projected to decrease by 1.5 ± 1.1 psu, and the liquid freshwater column is projected to increase by 5.4 ± 3.8 m by the end of the 21st century under the Representative Concentration Pathway 8.5 (RCP8.5) scenario. Most freshening would occur in the Arctic Ocean basins, that is, the Canada, Makarov, and Amundsen basins. Anomalies in freshwater flux from sea ice melt, Bering Strait inflow, net precipitation (P‐E), river runoff, and freshwater through the Barents Sea Opening (BSO) would contribute to Arctic Ocean freshening. CMIP5 historical and RCP8.5 experiments showed that the respective projected contributions from BSO freshwater flux, river runoff, P‐E, and Bering Strait inflow are about 6.4, 5.0, 2.7, and 2.2 times the contribution from sea ice melt averaged throughout the 21st century. Contributions from sea ice melt and Bering Strait inflow would increase and then decrease gradually, while those from BSO freshwater flux, river runoff, and P‐E would increase continuously. The CMIP5 models are able to simulate the Arctic Ocean freshwater system more accurately than CMIP3 models. However, the simulated rate of increase of freshwater content (296 ± 232 km3/yr) is weaker than estimated (600 ± 300 km3/yr) based on observations (1992–2012). Moreover, the simulated BSO and Davis Strait freshwater fluxes still exhibit substantial intermodel spread and they differ considerably from observed values.

Response of Arctic Freshwater to the Arctic Oscillation in Coupled Climate Models

“…The models show some common biases with respect to observations (e.g., the Polar Science Center Hydrographic Climatology, Steele et al 2001): the Beaufort gyre is more spatially diffuse, situated further from Canada and Alaska, and the central Arctic is overly fresh. These biases are also clear in the CMIP5 multi-model mean freshwater distribution (Shu et al 2018).…”

“…Ding et al (2016) examined the processes governing the seasonal freshwater cycle in CMIP5 models, but otherwise the only intercomparison of Arctic Ocean freshwater and/or hydrography in models of the same generation of the CMIP5 models thus far comes from Arctic Ocean CORE-II studies, which considered forced (non-coupled) model simulations (Wang et al 2016;Ilicak et al 2016). Shu et al (2018) assess the projected changes of the CMIP5 multi-model mean freshwater content in the 21st century.…”

“…For comparison, Shu et al (2018) show that the CMIP5 historical run multi-model mean captures c. 50% of the observed freshening during the 1992-2012 observational period. This signal can be assumed to average across stochastic SLP forcing and represent the effects of longer-term anthropogenic climate change, relating to sea-ice melt and an enhanced hydrological cycle, amongst other possible changes (which could include changes to SLP variability).…”

Understanding Regional and Seasonal Variability Is Key to Gaining a Pan-Arctic Perspective on Arctic Ocean Freshening

“…Pacific water inflows Increased inflows are projected to at least mid-twenty-first century (e.g, Shu et al, 2018), here the monthly projections are based on the observed mean monthly increases in PW over the last 25 years (cf. Woodgate, 2018vs.…”

Nordic Seas Acidification

“…In the Iceland Sea, above 200 m, there is, in contrast to the other regions, a decrease in ALK in the RCP8.5 simulation, which has a negative impact on the pH. This is probably related to an increased export of freshwater from the Arctic Ocean (Shu et al, 2018).…”